Hyperactivity Study

[jordan]

I was asked this weekend for psychology class to develop a study to test the effectiveness of a drug on children with hyperactivity. This study (and the drug) aren't real, but we must remain within the guidelines of the scienetific method and ethics.

 

Now, my plan is to have the children focus in a closed room on a monotonous task and see how long they keep working until one of a set group of actions indicative of distraction are noticed. Then, after taking the drug, the process will be repeated and I will have to look for a change in the length of time they remained focused.

 

So, the question: I need a control group. I can't seem to come up with what I would need a control group for in this study. I have considered using a placebo, but I would doubt a mental disorder can be cured by playing a mental trick on someone. If these children have no control over how long they can focus, does it make sense that telling them they will be cured with a drug wont give them any more control? If so, what can be used as a control?

[badchad]

You don't necessarily need a "placebo". You need to perform a "double blind" study. In this case, neither the person observing the students and collecting the data, nor the children themselves know what drug they've been given. This will eliminate bias. Then simply measure the response of the children to the two drugs (e.g. non-drug or placebo VS drug).

 

If you have a large enough number of children, you could also perform the test using three drugs or conditions, where you could split the hyper kids into three different groups: 1. Those receiving placebo, 2. Those receiving a drug which is already known to reduce hyperactivity, and 3. The "new" drug.

 

The latter situation is probably a bit more realistic. This is because there are already drugs used to treat hyperactivity. Since there are already drugs available, we only care whether a new drug is better then whats currently available.

[Glider]

So, the question: I need a control group. I can't seem to come up with what I would need a control group for in this study. I have considered using a placebo, but I would doubt a mental disorder can be cured by playing a mental trick on someone. If these children have no control over how long they can focus, does it make sense that telling them they will be cured with a drug wont give them any more control? If so, what can be used as a control?

 

Even double-blind studies need to be placebo controlled. The point of a placebo is that it is something that is known to have no effect at all. It's not supposed to 'cure' anything. Giving a placebo simply rules out the chance that the simple act of administration didn't have an effect.

 

When giving a placebo, you don't tell the participant it's going to have x, y or z effect. This would contaminate the experiment by introducing participant expectation as a factor (it may or may not be the case, but the point is you couldn't know which).

 

This also applies to the experimental intervention (experimental drug). You don't tell the participant what the drug is supposed to do, unless you are investigating the effects of your suggestion.

 

This introduces a problem with informed consent. However, it does mean that it's easier to avoid experimental contamination with children (< 16 yrs). This is because they are not in a position to provide informed consent. It has to be obtained from their parents. So, all the information that might contaminate the experiment is givent to the parents; what the drug is, what it is supposed to do and whether there are any possible side effects, etc..

[badchad]

Well, in certain instances you don't need a placebo control. Generally this occurs when there is already a known drug which is effective at treating a disorder.

 

An example would be the evaluation of antipsychotic drugs. The early drugs used to treat psychosis and schizophrenia were known as "atypical" antipsychotics. Since these drugs were efficacious in resolving symptoms newer drugs are compared against this standard. For instance, we could spend a lot of time and money doing a double blind study of a "new drug" VS. placebo. However, if the "new drug" isn't better then the old drug, we've wasted our time.

 

So, if you have a "new drug", you could in fact compare it only with an older standard drug, and not use a placebo. This would be because both groups receive a drug ("new drug" and old drug), thus, both would have an equal amount of placebo effect.

 

However, a double blind without a placebo would only occur when there is already a standard drug with which to compare it too.

[Glider]

As a rule, you'd find it hard to get your study published without a control. Whilst 'new drug' Vs 'existing drug' can be tested with two groups, a control is still needed to make sure no unexpected experimental confound influenced the results, and that your sample was not skewed in some way.

 

If your sample doesn't yield normative baseline results, then either your sample is not representative, or the experiment is contaminated. If your control group shows a difference across measures, then your experiment is contaminated.

 

The control group doesn't simply act as a neutral comparison, it acts as an experimental control (the clue's in the name).

[badchad]

This is an interesting question to me Glider, as I've been looking at a lot of clinical trials in preparation for a lecture on antipsychotic drugs.

 

I do agree that a 3 drug design (placebo VS. drug VS. New drug) is scientifically sound, and preferred.

 

However, after the initial efficacy of a drug has been demonstrated doing a 2 drug design (New drug VS. placebo); a 3 drug design is not as cost effective as a two drug design. For instance, in a two drug design you can determine whether the new drug is better then the old drug. Why would this need a placebo "control"? both groups receive a drug, so both would be subject to the same amount of "placebo effect". The same holds true for unexpected experimental confounds, they should be equal amongst the two treatments. Thus, any differences that arise would be due to the drug itself. In addition, you run into ethical problems that can occur from not treating a patient (those receiving "placebo"). I'm just curious as to people's input. I've listed a few two-drug designs if you're interested.

 

Am J Psychiatry. 2004 Jun;161(6):985-95.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=15169686

 

J Clin Psychiatry. 2004 May;65(5):696-701.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=15163258

 

Int Clin Psychopharmacol. 2004 Mar;19(2):63-9.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=15076013

 

Eur Neuropsychopharmacol. 1997 May;7(2):125-37.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=9169300

 

Am J Psychiatry. 1997 Apr;154(4):457-65.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=9090331

 

Ann Clin Psychiatry. 1993 Sep;5(3):199-207.

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=pubmed&dopt=Abstract&list_uids=8281243

 

These are just a few two drug designs. IMO it seems that a medline search yields MORE two drug designs then three drug designs. These are clinical trials, which are rigorously planned and well-thought out (generally speaking). In other words, it seems as though a two drug design is well accepted when comparing the effects of two drugs and ti wouldn't be difficult to publish a two drug design.

[Glider]

Where the participants are patients undergoing prescribed drug therapy, it would not be ethical to use a placebo control. That would mean taking patients off medication, or denying them medication. In these cases, two sample crossover designs are preferred. The crossover acts as a control.